betaFeedback

WorksEngineering

Three-Dimensional Impact Angle Control Guidance Law with Field-of-View Constraint

Abstract

This paper presents a novel impact angle control guidance law to address the challenge of achieving any desired impact angle while satisfying the seeker's Field Of View (FOV) limit and maintaining acceleration continuity. The guidance problem is decomposed into two two-dimensional (2D) engagement planes. In the first plane, a pure proportional navigation guidance (PNG) law is employed toward a virtual target to align the missile velocity vector with the line-of-sight (LOS) to the target. In the second plane, a two-stage time-varying biased PNG (TVBPN) strategy is utilized to accurately shape the trajectory and achieve the desired terminal impact angle while maintaining compliance with FOV limitations. The position of the virtual target point ensures a smooth and continuous lateral acceleration at the transition between the first and second stages. The effectiveness and feasibility of the proposed law were validated through numerical simulations.

NOMENCLATURE

aa Speed of sound

aMa_M Lateral acceleration command

bb Bias term

B(t)B(t) Bias integral

BrefB_{\mathrm{ref}} Reference value of the bias integral

CDC_D Total drag coefficient

CD0C_{D_0} Zero-lift drag coefficient

CLC_L Lift coefficient

DD Aerodynamic drag

eLOS\mathbf{e}_{LOS} Line-of-sight unit vector

em\mathbf{e}_m Missile velocity unit vector

eN\mathbf{e}_N Unit normal to the engagement plane

gg Gravitational acceleration

hh Altitude

kk Virtual target scaling factor

KK Induced drag coefficient

mm Mass

MM Mach number

NN Navigation gain

QQ Dynamic pressure

rr Relative range between missile and target

RR Universal gas constant

SS Reference area

tt Time

tst_s Switching instant

TT Thrust

TT Atmospheric temperature

VV Constant missile speed

VMV_M Time-varying missile velocity

Greek Symbols

γ\gamma Flight path angle / Heat capacity ratio

γL\gamma_L Critical impact angle

θ\theta Elevation angle

λ\lambda Line-of-sight (LOS) angle

ρ\rho Atmospheric density

σ\sigma Seeker look angle

σmax⁡\sigma_{\max} Maximum allowable look angle

τ1,τ2\tau_1,\tau_2 Time constants

ψ\psi Azimuth angle

INTRODUCTION

Effective missile guidance is pivotal for modern defence systems. Achieving a prescribed impact angle is crucial for maximizing warhead effectiveness and minimizing collateral damage. Since FOV (Field-of-View) constraints are inherent in any realistic seeker system, guidance laws are also required to ensure the target remains within the seeker’s FOV throughout the engagement. Moreover, discontinuous acceleration commands can induce excessive structural loads on the missile airframe. Therefore, maintaining a continuous acceleration profile is essential for practical implementation.

In recent years, numerous studies have been conducted to design of impact angle control guidance (IACG) laws for planar engagements. Ratnoo and Ghose [1] proposed a composite proportional navigation (PN)-based guidance law by combining an orientation phase with classical PN in the terminal phase to achieve any desired impact angle. This study was extended in [2] to address the interception of nonstationary, nonmaneuvering targets. In [3], FOV constraints were incorporated into a two-stage PN guidance (PNG) law while achieving all feasible impact angles. The same objectives were achieved with a reduced control effort using a three-staged PN guidance law by Ranjan et al. [4]. Since these guidance laws are based on multiple PN stages, the resulting missile lateral acceleration profiles are discontinuous and exhibit abrupt jumps. Park et al. [5, 6] designed a two-staged biased proportional navigation guidance (BPNG) law that considers FOV and lateral acceleration constraints. An energy-optimal impact angle control guidance law that explicitly incorporates FOV constraints for missiles equipped with strap-down seekers was presented in [7]. Lee et al. [8] proposed a time-to-go polynomial guidance (TPG) law while satisfying both terminal acceleration and seeker FOV constraints, where the acceleration command was formulated as a polynomial function of time-to-go. A time-varying two-staged BPNG law was designed by Yang et al. [9] to achieve all possible impact angles while maintaining an FOV constraint. To further circumvent structural complexities, Zhang and Wang [10] proposed a guidance law featuring a time-varying navigation gain to achieve specific terminal angles without requiring phased switches or the addition of bias terms. Additionally, Gao et al. [11] investigated an adaptive terminal guidance strategy specifically focused on satisfying rigorous impact-angle constraints. However, all aforementioned guidance laws are limited to two-dimensional engagement scenarios, which restricts their applicability to realistic three-dimensional (3D) interception problems.

To address the complexities of realistic missile-target interaction, recent research has extended impact angle control guidance into three-dimensional frameworks. Erer and Tekin [12] proposed guidance laws that extend biased PN to 3D engagement by rotating either the line of sight (LOS) or the velocity vector toward the desired terminal direction. In [13], a quaternion-based guidance law was proposed that extends a 2D BPNG law to 3D by incorporating quaternion rotations between the missile velocity vector and the desired impact direction. Wang et al. [14] presented a nonlinear guidance law that enables simultaneous control of impact angle and impact time using a two-stage structure, consisting of a sliding-mode-based IACG phase followed by a PNG phase. Another nonlinear guidance law, proposed in [15], ensures interception within a predefined convergence time, using a sliding mode control framework and a predefined-time stability theory. Nanavati et al. [16] proposed a nonlinear guidance law that uses a finite-time state-dependent Riccati equation and a lead-angle formulation. Zhou et al. [17] developed an IACG law by decomposing the trajectory into two independently designed nominal curves, an elliptical path in the vertical plane and a Bézier curve in the lateral plane.

Advanced nonlinear techniques, such as the fuzzy variable coefficients strategy combined with sliding mode control, have also been investigated by Xiong et al. [18] to enhance the robustness of 3D guidance profiles. Moreover, recent advances have successfully incorporated learning-based adaptive manoeuver guidance [19] and three-dimensional target manoeuver compensation algorithms [20] to tackle highly dynamic engagements. However, these 3D engagement studies do not explicitly consider the FOV constraints of the seeker.

In practical engagement scenarios, strict adherence to FOV constraints is indispensable due to the physical limitations of onboard seekers. Consequently, guidance laws must explicitly account for FOV limitations to ensure that the target remains within the seeker’s tracking capability throughout the engagement. Duvvuru et al. [21] proposed a guidance law that satisfies the terminal impact angle and the FOV constraints while maximizing the terminal speed and minimizing flight time, using an extended form of the generalized model predictive static programming technique. In [22], a guidance law was proposed that enforces prescribed performance and FOV constraints by shaping LOS angle errors through a range-dependent prescribed performance function. A predefined-time IACG law was introduced in [23] that uses a novel predefined-time error dynamic tailored for second-order systems and auxiliary functions to prevent FOV violations. Wang et al. [24] proposed an FOV-constrained guidance law that combines a PN term with a bias component shaped by a nonlinear error dynamic for speed-varying missiles. Another FOV-constrained guidance law [25] that constructs range profiles as cubic polynomials of LOS angles was proposed for speed-varying missiles. Addressing the challenge of diverse target kinematics, Du et al. [26] recently introduced a sliding mode guidance scheme to enforce field-of-view limits against targets displaying various motion profiles. Hu et al. [27] proposed a guidance law that achieves both impact angle and seeker FOV constraints using polynomial shaping of LOS angles in pitch and yaw, formulated as cubic polynomials of relative range, and a similar approach was reported in [28]. Ghosh et al. [29] proposed a composite PN-based guidance law (μαCIG3D\mu\alpha\mathrm{CIG3D}) that enables unmanned aerial vehicles to approach a stationary target from any arbitrary direction in 3D space while accounting for FOV limitations; however, this law results in multiple discontinuities in the lateral acceleration profile. Despite these developments, no existing 3D IACG framework simultaneously guarantees strict FOV constraint satisfaction, continuous acceleration profiles, and the ability to achieve arbitrary terminal impact directions.

Motivated by these gaps, this paper proposes a novel three-dimensional impact angle control guidance law for stationary targets that ensures strict FOV compliance while maintaining acceleration continuity. The key contributions are summarized as follows:

  • A novel 3D impact angle control guidance scheme is developed by decoupling the spatial engagement into two sequential planar phases, enabling target interception from any arbitrary spatial direction while strictly enforcing seeker FOV constraints.

  • An analytical virtual target placement strategy is introduced for the primary alignment phase to completely eliminate the discontinuous lateral acceleration transients typically associated with multi-stage guidance transitions.

  • A modified switching logic for the time-varying biased proportional navigation (TVBPN) law is formulated to ensure deterministic acceleration continuity and reliable phase initialization even for highly shallow terminal impact angles.

The remainder of this paper is organized as follows. The problem statement and a background review of planar guidance laws are established in Section 2. Section 3 details the formulation of the 3-staged 3D IACG scheme, including the modified TVBPN logic, virtual target placement, and analytical proof of FOV satisfaction. Numerical simulation results are presented in Section 4, and Section 5 concludes the paper.

PROBLEM STATEMENT AND BACKGROUND

Problem Statement

Consider an inertial coordinate frame (X,Y,Z)(X,Y,Z) centered at an origin OO. Let MM and TT denote the initial spatial coordinates of the missile and a stationary target, respectively, as shown in Figure 1. The three-dimensional engagement is decoupled into two sequential planar phases. The first plane is defined by the initial line-of-sight (LOS) unit vector, eLOS1\mathbf{e}_{\mathrm{LOS}_1}, and the initial missile velocity unit vector, em0\mathbf{e}_{m0}. The missile targets a virtual point T′T' along the initial LOS until its velocity vector aligns with the LOS at the transition point P12P_{12}. The second plane is formed by the LOS unit vector at transition, eLOS2\mathbf{e}_{\mathrm{LOS}_2}, and the desired terminal impact unit vector, emf\mathbf{e}_{mf}. The unit normals to the first and second engagement planes are respectively defined as:

Figure 1. 3D Engagement Scenario

eN1=eLOS1×em0,eN2=emf×eLOS2(1)\begin{aligned} \mathbf{e}_{N_1} &= \mathbf{e}_{\mathrm{LOS}_1} \times\mathbf{e}_{m0},\\ \mathbf{e}_{N_2} &= \mathbf{e}_{mf} \times\mathbf{e}_{\mathrm{LOS}_2} \tag*{(1)} \end{aligned}

Local 2D Cartesian frames are established for each plane to formulate the guidance problem. The first frame originates at MM, with the principal axes aligned along eLOS1\mathbf{e}_{\mathrm{LOS}_1} and eN1×eLOS1\mathbf{e}_{N_1} \times\mathbf{e}_{\mathrm{LOS}_1}. The second frame originates at the switching point P12P_{12}, with axes along eLOS2\mathbf{e}_{\mathrm{LOS}_2} and eN2×eLOS2\mathbf{e}_{N_2} \times\mathbf{e}_{\mathrm{LOS}_2}.

Since the three-dimensional engagement is decomposed into sequential planar sub-engagements, the governing dynamics within each phase can be rigorously described using standard planar interception kinematics, as shown in Figure 2. Let VV denote the constant missile speed, rr the relative range between the missile and the target, γ\gamma the flight path angle, λ\lambda the line-of-sight (LOS) angle, and σ=γ−λ\sigma= \gamma- \lambda the seeker look angle. The corresponding planar kinematic relations are given by:

Figure 2. 2D Engagement Geometry

r˙=−Vcos⁡σ,λ˙=−Vsin⁡σr,γ˙=aMV(2)\begin{aligned} \dot{r} &= -V\cos{\sigma},\\ \dot{\lambda} &= -\frac{V\sin{\sigma}}{r},\\ \dot{\gamma} &= \frac{a_M}{V} \tag*{(2)} \end{aligned}

Background

For a pure proportional navigation (PN) guidance law with navigation gain NN, the lateral acceleration command aMa_M is given by

aM=NVλ˙(3)a_M = NV\dot{\lambda} \tag*{(3)}

A biased proportional navigation (BPN) guidance law incorporates an additive bias term bb to shape the trajectory, yielding the following lateral acceleration command:

aM=NVλ˙+Vb(4)a_M = NV\dot{\lambda} + Vb \tag*{(4)}

Using (2), the flight path angle rate is expressed as

γ˙=Nλ˙+b(5)\dot{\gamma} = N\dot{\lambda} + b \tag*{(5)}

Integrating (5) over the engagement interval t∈[0,tf]t \in[0,t_f] yields the terminal flight path angle:

γf−γ0=N(λf−λ0)+∫0tfb dt(6)\gamma_f - \gamma_0 = N(\lambda_f - \lambda_0) + \int_{0}^{t_f} b\,dt \tag*{(6)}

For successful interception, the terminal condition γf=λf\gamma_f = \lambda_f must be satisfied. Enforcing this constraint yields the required reference value of the bias integral, BrefB_{\mathrm{ref}}:

Bref=∫0tfb dt=(1−N)γf+Nλ0−γ0(7)B_{\mathrm{ref}} = \int_{0}^{t_f} b\,\mathrm{d}t = (1-N)\gamma_f + N\lambda_0 - \gamma_0 \tag*{(7)}

Review of the TVBPN Guidance Law

To address 2D FOV-constrained IACG problems, the time-varying biased proportional navigation (TVBPN) law (9) utilizes two distinct time-varying bias terms, b1b_1 and b2b_2, governed by a switching logic. During the initial phase, the bias b1b_1 drives the look angle σ\sigma toward its allowable limit σmax⁡\sigma_{\max}:

b1=(1−N)λ˙+1τ1(σmax⁡−σ)(8)b_1 = (1-N)\dot{\lambda} + \frac{1}{\tau_1}(\sigma_{\max}-\sigma) \tag*{(8)}

where, τ1\tau_1 is the time constant. Substituting b1b_1 into the BPN acceleration command yields:

am=Vλ˙+Vτ1(σmax⁡−σ)(9)a_m = V\dot{\lambda} + \frac{V}{\tau_1}(\sigma_{\max}-\sigma) \tag*{(9)}

The subsequent phase utilizes b2b_2 to drive the bias integral B(t)B(t) to the required reference value BrefB_{\mathrm{ref}} necessary for the desired terminal impact angle:

b2=B˙(t)=Bref−B(t)τ2(10)b_2 = \dot{B}(t) = \frac{B_{\mathrm{ref}}-B(t)}{\tau_2} \tag*{(10)}
⇒ b2=(Bref−B(ts))exp⁡(−t−tsτ2)τ2(11)\Rightarrow\ b_2 = \frac{\left(B_{\mathrm{ref}}-B(t_s)\right)\exp\left(-\frac{t-t_s}{\tau_2}\right)}{\tau_2} \tag*{(11)}

where, τ2\tau_2 is the convergence time constant and tst_s denotes the switching instant. The corresponding lateral acceleration command is given by

aM=Nλ˙+Bref−B(t)τ2(12)a_M = N\dot{\lambda} + \frac{B_{\mathrm{ref}}-B(t)}{\tau_2} \tag*{(12)}

The transition between the two bias formulations is governed by their relative magnitudes, with the switching logic defined as:

b={b1,∣b1∣<∣b2∣b2,∣b1∣≥∣b2∣(13)b = \begin{cases} b_1, & |b_1| < |b_2| \\ b_2, & |b_1| \ge|b_2| \end{cases} \tag*{(13)}

3-STAGED 3D IACG LAW

The proposed 3D IACG scheme guarantees prescribed terminal impact angles, strict seeker FOV compliance, and continuous lateral acceleration. The three-dimensional engagement is decoupled into a pure PNG phase on the primary plane, followed by a two-stage time-varying BPN (TVBPN) phase on the secondary plane. To eliminate lateral acceleration discontinuities during the planar transition, a virtual target is strategically positioned on the initial LOS.

Design of the 3-staged 3D IACG law

The missile operates at a constant speed VV, with control authority exerted exclusively via lateral acceleration. During the first stage on the primary plane (X1,Y1)(X_1, Y_1), a pure PNG law with a navigation gain of N=2N = 2 is applied to drive the missile toward a virtual target located at TT on the initial LOS, as shown in Figure 1. The corresponding acceleration command and flight path angle rate are given by:

aM1=2Vλ˙1,γ˙1=2λ˙1(14)a_{M_1} = 2V\dot{\lambda}_1, \\ \dot{\gamma}_1 = 2\dot{\lambda}_1 \tag*{(14)}

where, the subscript ‘1’ denotes variables in the primary plane. For N=2N = 2, the look angle rate reduces to σ˙1=λ˙1\dot{\sigma}_1 = \dot{\lambda}_1. This phase continues until the missile velocity vector aligns with the line-of-sight (LOS) direction in the inertial frame (i.e., γ=λ\gamma= \lambda). Based on the frame definitions, the initial conditions are given by:

σ10=γ10,λ10=0(15)\sigma_{10} = \gamma_{10}, \\ \lambda_{10} = 0 \tag*{(15)}

Upon alignment of the velocity vector with the LOS, the engagement transitions to the secondary plane (X2,Y2)(X_2, Y_2), where the initial conditions reset relative to the new LOS:

σ20=0,λ20=0,γ20=0(16)\sigma_{20} = 0, \\ \lambda_{20} = 0, \\ \gamma_{20} = 0 \tag*{(16)}

The lateral acceleration and the flight path angle rate in the secondary plane are governed by an augmented TVBPN law with navigation gain N2N_2:

aM2=N2Vλ˙2+Vb,γ˙2=N2λ˙2+b(17)a_{M_2} = N_2V\dot{\lambda}_2 + Vb, \\ \dot{\gamma}_2 = N_2\dot{\lambda}_2 + b \tag*{(17)}

This yields a look angle rate of σ˙2=(N2−1)λ˙2+b\dot{\sigma}_2 = (N_2 - 1)\dot{\lambda}_2 + b. Here, the subscript ‘2’ denotes variables associated with the secondary plane.

Modified TVBPN Switching Logic

The standard TVBPN guidance logic assumes ∣b1∣<∣b2∣|b_1| < |b_2| initially. However, for small desired terminal impact angles (typically <5∘< 5^\circ), the required bias integral BrefB_{\mathrm{ref}} becomes sufficiently small, causing ∣b2∣<∣b1∣|b_2| < |b_1| at the onset of the secondary plane. If b2b_2 is selected initially, enforcing acceleration continuity during the planar transition becomes analytically intractable, since b2b_2 depends explicitly on the terminal condition.

To address this limitation, the switching logic is modified to ensure b1b_1 is strictly enforced as the initial bias of the secondary phase. Define a critical impact angle γL\gamma_L such that the initial bias magnitudes are equal, i.e., ∣b10∣=∣b20∣|b_{10}| = |b_{20}|. Using the initial conditions in (16), along with (8) and (10), this condition reduces to:

∣(1−N2)λ2+1τ1(σmax⁡−σ2)∣=∣Bref−B(t)τ2∣⇒σmax⁡τ1=(1−N2)γLτ2⇒γL=σmax⁡1−N2τ2τ1(18)\left|(1-N_{2})\lambda_{2}+\frac{1}{\tau_{1}}(\sigma_{\max}-\sigma_{2})\right|=\left|\frac{B_{\mathrm{ref}}-B(t)}{\tau_{2}}\right| \\ \Rightarrow\quad\frac{\sigma_{\max}}{\tau_{1}}=\frac{(1-N_{2})\gamma_{L}}{\tau_{2}} \\ \Rightarrow\quad\gamma_{L}=\frac{\sigma_{\max}}{1-N_{2}}\frac{\tau_{2}}{\tau_{1}} \tag*{(18)}

The switching criterion is subsequently modified into a piecewise condition based on the desired terminal angle γf\gamma_f:

b={b1,∣γf∣<∣γL∣,{b1,∣b1∣<∣b2∣,b2,∣b1∣≥∣b2∣,∣γf∣≥∣γL∣.(19)b=\begin{cases} b_{1}, & |\gamma_f|<|\gamma_L|,\\ \begin{cases} b_{1}, & |b_{1}|<|b_{2}|,\\ b_{2}, & |b_{1}|\ge|b_{2}|, \end{cases} & |\gamma_f|\ge|\gamma_L|. \end{cases} \tag*{(19)}

This structured modification guarantees that b1b_1 initiates the secondary phase, thereby enabling deterministic enforcement of acceleration continuity while maintaining compliance with the FOV constraint.

Virtual Target Placement

The position of the virtual target is calculated by enforcing continuity of the lateral acceleration magnitude across the planar transition, thereby avoiding structural loading transients:

∣aM1(ts1)∣=∣aM2(0)∣(20)|a_{M_1}(t_{s_1})|=|a_{M_2}(0)| \tag*{(20)}

where, ts1t_{s_1} denotes the transition time from the primary to the secondary plane. Substituting the respective acceleration commands yields:

∣2Vλ˙1(ts1)∣=∣N2Vλ˙2(0)+Vb1(0)∣(21)|2V\dot{\lambda}_1(t_{s_1})|=|N_2V\dot{\lambda}_2(0)+Vb_1(0)| \tag*{(21)}

For a pure PNG law with N=2N=2 against a stationary target, the missile follows a circular trajectory, resulting in constant lateral acceleration. Consequently, the acceleration at the transition instant equals its initial value in the primary phase. By defining the virtual target T′T' at a fractional distance kk along the initial relative range r0r_0, the continuity condition simplifies to:

2V2sin⁡(γl0)kr0=Vσmax⁡τ1(22)\frac{2V^2\sin(\gamma_{l0})}{kr_0}=\frac{V\sigma_{\max}}{\tau_1} \tag*{(22)}

Solving for the virtual target scaling factor kk yields:

k=2Vτ1sin⁡(γl0)r0σmax⁡(23)k=\frac{2V\tau_1\sin(\gamma_{l0})}{r_0\sigma_{\max}} \tag*{(23)}

Satisfaction of the FOV limit

To ensure continuous target tracking, the seeker look angle must remain strictly bounded throughout the engagement. In the primary phase, a pure PNG law with N=2N=2 is employed. As established in (3), for N>1N>1, the look angle dynamics satisfy:

σ˙1={>0∀−π<σ1<0, N>1<0∀0<σ1<π, N>1=0if σ1=0, π=0if N=1(24)\dot{\sigma}_1 = \begin{cases} >0 & \forall-\pi< \sigma_1 < 0,\ N > 1 \\ <0 & \forall0 < \sigma_1 < \pi,\ N > 1 \\ =0 & \text{if } \sigma_1 = 0,\ \pi\\ =0 & \text{if } N = 1 \end{cases} \tag*{(24)}

i.e., for any N>1N > 1 and positive initial look angle, the look angle rate is strictly negative (σ˙1<0\dot{\sigma}_1 < 0), ensuring σ1\sigma_1 monotonically approaches zero.

Upon transition to the secondary plane, the look angle dynamics are governed by the modified TVBPN law. During the b1b_1 phase:

σ˙2(t)=(N2−1)λ˙2+b1=1τ1(σmax⁡−σ2)(25)\dot{\sigma}_2(t) = (N_2 - 1)\dot{\lambda}_2 + b_1 = \frac{1}{\tau_1}(\sigma_{\max} - \sigma_2) \tag*{(25)}

which implies exponential convergence of σ2\sigma_2 toward the upper bound σmax⁡\sigma_{\max}. Once the switching condition activates the b2b_2 phase at time ts2t_{s_2}, the look angle rate becomes:

σ˙2(t)=(N2−1)λ˙2+b2=(1−N2)Vsin⁡σ2(t)r(t)+(Bref−B(ts2))exp⁡(−(t−ts2)/τ2)τ2(26)\dot{\sigma}_2(t) = (N_2 - 1)\dot{\lambda}_2 + b_2 = \frac{(1-N_2)V\sin\sigma_2(t)}{r(t)} + \frac{(B_{\mathrm{ref}} - B(t_{s_2}))\exp(-(t-t_{s_2})/\tau_2)}{\tau_2} \tag*{(26)}

As shown in (9), this condition ensures that σ˙2(t)≤0\dot{\sigma}_2(t) \leq0 for t≥ts2t \geq t_{s_2}, implying that σ2\sigma_2 remains bounded and does not exceed σmax⁡\sigma_{\max}. Therefore, the seeker FOV constraint is satisfied throughout all phases of the engagement.

NUMERICAL ANALYSIS

The performance of the proposed 3D IACG scheme is evaluated through numerical simulations. Unless otherwise specified, the missile and target are initially located at [0,0,0]T[0,0,0]^T m and [2500,2500,2500]T[2500,2500,2500]^T m, respectively. The missile maintains a constant speed of 50 m/s. The maximum seeker FOV is constrained by a maximum allowable look angle of σmax⁡=45∘\sigma_{\max} = 45^\circ. The guidance parameters are set to N2=3N_2 = 3, τ1=2\tau_1 = 2 s, and τ2=0.5\tau_2 = 0.5 s. The initial velocity vector is chosen such that the initial look angle satisfies σ0<σmax⁡\sigma_0 < \sigma_{\max}. The spatial orientation of an arbitrary vector A⃗=[ax,ay,az]T\vec{A} = [a_x,a_y,a_z]^T is parameterized by its azimuth angle ψ∈[−π,π]\psi\in[-\pi,\pi] and elevation angle θ∈[−π/2,π/2]\theta\in[-\pi/2,\pi/2], as illustrated in Figure 3, and defined as:

Figure 3. 3D Angles

ψ=tan⁡−1(ayax),θ=tan⁡−1(azax2+ay2)(27)\begin{aligned} \psi&= \tan^{-1}\left(\frac{a_y}{a_x}\right), \\ \theta&= \tan^{-1}\left(\frac{a_z}{\sqrt{a_x^2+a_y^2}}\right) \tag*{(27)} \end{aligned}

Results with Varying Impact Angles

To demonstrate terminal angle control authority, the missile is launched with (ψ0,θ0):(45∘,0∘)(\psi_0,\theta_0): (45^\circ,0^\circ) and commanded to intercept the target at five distinct spatial orientations (ψf,θf):(−45∘,70∘)(\psi_f,\theta_f): (-45^\circ,70^\circ), (−60∘,−30∘)(-60^\circ,-30^\circ), (−30∘,−60∘)(-30^\circ,-60^\circ), (−90∘,45∘)(-90^\circ,45^\circ), and (−179∘,−75∘)(-179^\circ,-75^\circ).

As depicted in Figure 4a, successful interception is achieved for all prescribed terminal angles. The lateral acceleration profiles in Figure 4b remain perfectly continuous across all phase transitions, validating the virtual target placement strategy. The seeker look angle in Figure 4c exhibits the designed three-stage behavior: monotonically decaying to zero in the primary plane, exponentially approaching σmax⁡\sigma_{\max} in the secondary plane, and finally converging to zero upon interception. Notably, the look angle remains strictly bounded by the 45∘45^\circ limit throughout the engagement. Figure 4d and Figure 4e show the azimuth and elevation angle histories corresponding to the different terminal impact angle cases.

Figure 4. Simulation results of 3-Staged 3D IACG Law - Varying Impact Angles

Results with Varying Launch Angles

This scenario evaluates the robustness of the scheme against varying initial heading errors. The missile is launched from three distinct initial orientations: (ψ0,θ0):(20∘,70∘)(\psi_0, \theta_0) : (20^\circ, 70^\circ), (60∘,30∘)(60^\circ, 30^\circ), and (80∘,10∘)(80^\circ, 10^\circ), with terminal impact angles (ψf,θf):(−45∘,35.26∘)(\psi_f, \theta_f): (-45^\circ, 35.26^\circ).

Figure 5 confirms that the missile successfully intercepts the stationary target for all initial conditions. As the terminal constraints dominate the secondary planar engagement, the spatial trajectories in Figure 5a nearly coincide after the initial alignment phase. Acceleration continuity is preserved, as shown in Figure 5b, and the seeker FOV constraint is strictly satisfied, with the look angle remaining below σmax⁡\sigma_{\max}, as illustrated in Figure 5c. Furthermore, Figure 5d and Figure 5e demonstrate that the desired terminal impact angles are consistently achieved regardless of the initial launch orientation.

Figure 5. Simulation results of 3-Staged 3D IACG Law - Varying Launch Angles

Comparison Study

The proposed formulation is benchmarked against the composite PN-based μαCIG3D\mu\alpha\mathrm{CIG3D} law (29). For a representative engagement, the target is located at [2500,5000,5000]T m[2500, 5000, 5000]^{T}\,\mathrm{m}, and the missile is initialized with a velocity direction corresponding to (ψ0,θ0)=(45∘,35.26∘)(\psi_0, \theta_0) = (45^\circ, 35.26^\circ), with a commanded terminal impact angle of (ψf,θf)=(−45∘,−45∘)(\psi_f, \theta_f) = (-45^\circ, -45^\circ). While both guidance laws achieve successful interception, as shown in Figure 6a, the μαCIG3D\mu\alpha\mathrm{CIG3D} formulation exhibits multiple severe discontinuities in the lateral acceleration command, as illustrated in Figure 6b. In practical implementations, such abrupt variations would induce undesirable structural

Figure 6. (a) Trajectory; (b) Acceleration; (c) Look Angle; (d) Azimuth Angle; (e) Elevation Angle. Comparative results

transients in the airframe. In contrast, the proposed 3D IACG law yields a smooth and continuous acceleration profile. Furthermore, Figure 6d and Figure 6e demonstrate that the desired terminal impact angles are consistently achieved by both guidance laws.

To quantify efficiency, the total control effort (evaluated as ∫aM2 dt\int a_M^2\,dt) is compared across six diverse terminal impact geometries. As summarized in Table 1, the proposed scheme consistently demands lower total control effort than μα\mu\alphaCIG3D across all evaluated scenarios, highlighting the energetic efficiency of the continuous TVBPN transition over discrete composite formulations.

Methodψ\psi (deg)ϕ\phi (deg)Total control effort (m2^2/s3^3)
3-Staged 3D IACG-45-451928.40
μα\mu\alpha CIG3D-45-452054.20
3-Staged 3D IACG-60-301954.08
μα\mu\alpha CIG3D-60-302099.55
3-Staged 3D IACG-30-601923.86
μα\mu\alpha CIG3D-30-602044.83
3-Staged 3D IACG-90-452173.96
μα\mu\alpha CIG3D-90-452491.97
3-Staged 3D IACG-135-602299.14
μα\mu\alpha CIG3D-135-602704.93
3-Staged 3D IACG-180-752089.15
μα\mu\alpha CIG3D-180-752326.88

Table 1. Comparison of Total Control Effort for Different Initial Angle Cases

Realistic Simulations

To verify practical feasibility, the guidance scheme is applied to a realistic 6-DOF-equivalent point-mass missile model incorporating aerodynamic drag, gravity, and thrust profile variations. The governing equations for the time-varying velocity VMV_M, atmospheric density, and aerodynamic coefficients are detailed in the Appendix, adopted from (30). The missile is launched at 300 m/s along (ψ0,θ0):(45∘,0∘)(\psi_0,\theta_0):(45^\circ,0^\circ) toward a target at [2500,2500,2500]T[2500,2500,2500]^{T} m. Three distinct terminal impact angle cases are considered to evaluate the guidance performance, defined by (ψf,θf):(−60∘,30∘)(\psi_f,\theta_f):(-60^\circ,30^\circ), (−30∘,−60∘)(-30^\circ,-60^\circ), and (−179∘,−75∘)(-179^\circ,-75^\circ).

Figure 7a shows the trajectories for three different cases, while the corresponding lateral acceleration profiles are presented in Figure 7b. Despite the highly nonlinear velocity variation induced by aerodynamic drag and gravity, as shown in Figure 7c, the guidance scheme successfully enforces the prescribed terminal impact angles, as illustrated in Figure 7e and Figure 7f. Crucially, the seeker look angle, shown in Figure 7d, remains bounded below the 45° FOV limit. Although the lateral acceleration profiles in Figure 7b exhibit minor deviations from the ideal constant profile during the primary phase due to time-varying speed, the transition behavior remains smooth, confirming the robustness of the proposed scheme under realistic engagement conditions.

Trajectory

Figure 7. (a) Trajectory

Figure 7b. (b) Acceleration

Figure 7c. (c) Velocity

Look Angle

Figure 7d. (d) Look Angle

Figure 7e. (e) Azimuth Angle

Figure 7f. (f) Elevation Angle

Figure 7: Realistic Simulation results

Conclusions

A novel three-dimensional impact angle control guidance (IACG) law is developed to achieve specified terminal interception geometries while strictly enforcing seeker field-of-view (FOV) limits and guaranteeing lateral acceleration continuity. By decoupling the spatial engagement into two sequential planar phases, the scheme first employs a pure proportional navigation phase to align the velocity vector of the missile with the line-of-sight. The engagement then transitions to a secondary plane, where a modified two-stage time-varying biased proportional navigation (TVBPN) law actively shapes the terminal trajectory to satisfy the prescribed impact angle without violating FOV constraints. To eliminate the acceleration discontinuities typically associated with multi-stage guidance, an analytically positioned virtual target is introduced during the primary phase. Extensive nonlinear simulations, including comparative analysis and realistic missile dynamics, demonstrate the energetic efficiency, robustness, and practical feasibility of the proposed formulation.

Appendix

Realistic Vehicle Model Parameters

The 6-DOF-equivalent point-mass kinematics and aerodynamic properties utilized for the realistic simulations in Section 4.D are governed by the following equations, adopted from [30]. The time-varying missile velocity, VMV_M, is determined by the thrust TT, aerodynamic drag DD, mass mm, and elevation angle θ\theta:

V˙M=T−Dm−gsin⁡θ(28)\dot{V}_M = \frac{T-D}{m} - g\sin\theta \tag*{(28)}

where, gg is the gravitational acceleration. The aerodynamic drag DD and required lift coefficient CLC_L to sustain the commanded lateral acceleration aMa_M are formulated as:

D=QCDS,CL=maMQS(29)D = QC_D S,\qquad C_L = \frac{m a_M}{QS} \tag*{(29)}

where, S=1 m2S = 1\ \mathrm{m}^2 is the reference area, CDC_D is the total drag coefficient, and the dynamic pressure QQ is given by Q=12ρVM2Q = \frac{1}{2}\rho V_M^2. The total drag coefficient is a function of the zero-lift drag coefficient CD0C_{D0} and the induced drag coefficient KK:

CD=CD0+KCL2(30)C_D = C_{D0} + K C_L^2 \tag*{(30)}

The atmospheric temperature TT (in Kelvin), speed of sound aa, Mach number MM, and atmospheric density ρ\rho (in kg/m3\mathrm{kg}/\mathrm{m}^3) are modeled as functions of the altitude hh (in meters):

T={(288.16−0.0065h) Kh<11000 m216.66 Kh≥11000(31)T = \begin{cases} (288.16 - 0.0065h)\ \mathrm{K} & h < 11000\ \mathrm{m} \\ 216.66\ \mathrm{K} & h \ge11000 \end{cases} \tag*{(31)}
a=γRT,M=VMa(32)a = \sqrt{\gamma R T}, \qquad M = \frac{V_M}{a} \tag*{(32)}
ρ(h)=1.15579−1.058×10−4h+3.725×10−9h2−6.0×10−14h3(33)\rho(h) = 1.15579 - 1.058 \times10^{-4}h + 3.725 \times10^{-9}h^2 - 6.0 \times10^{-14}h^3 \tag*{(33)}

where, γ=1.4\gamma= 1.4 is the heat capacity ratio and R=287.05R = 287.05 J/(kg · K) is the universal gas constant. The aerodynamic coefficients CD0C_{D_0} and KK vary piecewise with the Mach number:

CD0={0.02M<0.930.02+0.2(M−0.93)M<1.030.04+0.06(M−1.03)M<1.100.0442−0.007(M−1.10)M≥1.10(34)C_{D_0} = \begin{cases} 0.02 & M < 0.93 \\ 0.02 + 0.2(M - 0.93) & M < 1.03 \\ 0.04 + 0.06(M - 1.03) & M < 1.10 \\ 0.0442 - 0.007(M - 1.10) & M \ge1.10 \end{cases} \tag*{(34)}
K={0.2M<1.150.2+0.246(M−1.15)M≥1.15(35)K = \begin{cases} 0.2 & M < 1.15 \\ 0.2 + 0.246(M - 1.15) & M \ge1.15 \end{cases} \tag*{(35)}

Finally, the thrust profile TT (in Newtons) of the vehicle and mass mm (in kg) are modeled as functions of the flight time tt (in seconds):

T={330000≤t≤1.575001.5≤t≤8.508.5≤t(36)T = \begin{cases} 33000 & 0 \le t \le1.5 \\ 7500 & 1.5 \le t \le8.5 \\ 0 & 8.5 \le t \end{cases} \tag*{(36)}
m={135−14.53t0≤t≤1.5113.205−3.31t1.5≤t≤8.590.0358.5≤t(37)m = \begin{cases} 135 - 14.53t & 0 \le t \le1.5 \\ 113.205 - 3.31t & 1.5 \le t \le8.5 \\ 90.035 & 8.5 \le t \end{cases} \tag*{(37)}

where, tt is the time from the launch.

References

  1. [1]Ratnoo, A. and Ghose, D. Impact angle constrained interception of stationary targets, Journal of Guidance, Control, and Dynamics, 2008, 31, (6), pp 1817–1822. doi: 10.2514/1.37864.DOI
  2. [2]Ratnoo, A. and Ghose, D. Impact angle constrained guidance against nonstationary nonmaneuvering targets, Journal of Guidance, Control, and Dynamics, 2010, 33, (1), pp 269–275. doi: 10.2514/1.45026.DOI
  3. [3]Ratnoo, A. Analysis of two-stage proportional navigation with heading constraints, Journal of Guidance, Control, and Dynamics, 2016, 39, (1), pp 156–164. doi: 10.2514/1.G001262.DOI
  4. [4]Ranjan, A., Hota, S. and Singh, N.K. Three-stage proportional navigation for intercepting stationary targets with impact angle constraints, Proceedings of the Sixth Indian Control Conference (ICC), 2019, pp 379–384. doi: 10.1109/ICC47138.2019.9123163.
  5. [5]Kim, T.-H., Park, B.-G. and Tahk, M.-J. Bias-shaping method for biased proportional navigation with terminal-angle constraint, Journal of Guidance, Control, and Dynamics, 2013, 36, (6), pp 1810–1816. doi: 10.2514/1.59252.DOI
  6. [6]Park, B.-G., Kim, T.-H. and Tahk, M.-J. Biased PNG with terminal-angle constraint for intercepting nonmaneuvering targets under physical constraints, IEEE Transactions on Aerospace and Electronic Systems, 2017, 53, (3), pp 1562–1572. doi: 10.1109/TAES.2017.2676518.DOI
  7. [7]Park, B.-G., Kim, T.-H. and Tahk, M.-J. Optimal impact angle control guidance law considering the seeker’s field-of-view limits, Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 2013, 227, (8), pp 1347–1364. doi: 10.1177/0954410012452367.DOI
  8. [8]Lee, C.-H., Kim, T.-H., Tahk, M.-J. and Whang, I.-H. Polynomial guidance laws considering terminal impact angle and acceleration constraints, IEEE Transactions on Aerospace and Electronic Systems, 2013, 49, (1), pp 74–92. doi: 10.1109/TAES.2013.6404092.DOI
  9. [9]9*. Yang, Z., Wang, H. and Lin, D. Time-varying biased proportional guidance with seeker’s field-of-view limit, International Journal of Aerospace Engineering, 2016, 2016. doi: 10.1155/2016/9272019.DOI
  10. [10]ZHANG, W.J. and WANG, B.M. A new guidance law for impact angle constraints with time-varying navigation gain, The Aeronautical Journal, 2022. doi: 10.1017/aer.2021.116.
  11. [11]GAO, C., LI, J., FENG, T. and JING, W. Adaptive terminal guidance law with impact-angle constraint, The Aeronautical Journal, 2017. doi: 10.1017/aer.2017.30.
  12. [12]ERER, K.S. and TEKIN, R. Impact vector guidance, Journal of Guidance, Control, and Dynamics, 2021, 44, (10), pp 1892–1901. doi: 10.2514/1.G006087.DOI
  13. [13]KIM, J.-H., PARK, S.-S., PARK, K.-K. and RYOO, C.-K. Quaternion based three-dimensional impact angle control guidance law, IEEE Transactions on Aerospace and Electronic Systems, 2021, 57, (4), pp 2311–2323. doi: 10.1109/TAES.2021.3058398.DOI
  14. [14]WANG, C., YU, H., DONG, W. and WANG, J. Three-dimensional impact angle and time control guidance law based on two-stage strategy, IEEE Transactions on Aerospace and Electronic Systems, 2022, 58, (6), pp 5361–5372. doi: 10.1109/TAES.2022.3169124.DOI
  15. [15]MAJUMDER, K. and KUMAR, S.R. Three-dimensional impact angle constrained nonlinear guidance with predefined convergence time, Nonlinear Dynamics, 2024, 112, (12), pp 9983–10008. doi: 10.1007/s11071-024-09564-1.DOI
  16. [16]NANAVATI, R.V., KUMAR, S.R. and MAITY, A. Lead-angle-based three-dimensional guidance for angle-constrained interception, Journal of Guidance, Control, and Dynamics, 2021, 44, (1), pp 190–199. doi: 10.2514/1.G005454.DOI
  17. [17]ZHOU, H., CHENG, T., LIU, X. and CHEN, W. Three-dimensional geometric descent guidance with impact angle constraint, IEEE Access, 2020, 8, pp 64932–64948. doi: 10.1109/ACCESS.2020.2984353.DOI
  18. [18]XIONG, J.H., TANG, S.J., GUO, J. and WANG, T.N. Improved sliding mode guidance law based on fuzzy variable coefficients strategy, The Aeronautical Journal, 2016. doi: 10.1017/aer.2015.42.
  19. [19]WU, T. and WANG, Z. Reinforcement learning-based adaptive spiral-diving manoeuver guidance method for reentry vehicles subject to unknown disturbances, The Aeronautical Journal, 2024. doi: 10.1017/aer.2024.18.
  20. [20]QI, G.Y., ZHANG, X.L., LI, L.Y., WANG, S.S. and ZHANG, H.Y. Integration of CFO-based three-dimensional target manoeuver compensation guidance rate and transition region optimisation algorithm, The Aeronautical Journal, 2025. doi: 10.1017/aer.2024.156.DOI
  21. [21]DUVVORU, R., MAITY, A. and UMAKANT, J. Three-dimensional field of view and impact angle constrained guidance with terminal speed maximization, Aerospace Science and Technology, 2022, 126, p 107552. doi: 10.1016/j.ast.2022.107552.
  22. [22]GU, Z., KANG, H. and SONG, S. Three-dimensional prescribed performance guidance law with field-of-view constraint and feasibility analysis, International Journal of Aeronautical and Space Sciences, 2024, 25, (4), pp 1536–1553. doi: 10.1007/s42405-024-00745-1.DOI
  23. [23]WANG, J., TAO, X., DONG, W. and WANG, C. Three-dimensional predefined-time impact angle control guidance law with field-of-view limit, Journal of the Franklin Institute, 2023, 360, (12), pp 7621–7644. doi: 10.1016/j.jfranklin.2023.06.021.DOI
  24. [24]WANG, J., DING, X., CHEN, Y., WANG, C. and XIN, M. Field-of-view constrained three-dimensional impact angle control guidance for speed-varying missiles, IEEE Transactions on Aerospace and Electronic Systems, 2022, 58, (5), pp 3992–4003. doi: 10.1109/TAES.2022.3157807.
  25. [25]Han, T., Hu, Q. and Xin, M. Three-dimensional approach angle guidance under varying velocity and field-of-view limit without using line-of-sight rate, IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2022, 52, (11), pp 7148–7159. doi: 10.1109/TSMC.2022.3150299.DOI
  26. [26]Du, H., Yang, M., Wang, S. and Chao, T. Impact angle control with field-of-view constrained guidance law against a various motion of target, The Aeronautical Journal, 2026. doi: 10.1017/aer.2025.12.
  27. [27]Hu, Q., Han, T. and Xin, M. Analytical solution for nonlinear three-dimensional guidance with impact angle and field-of-view constraints, IEEE Transactions on Industrial Electronics, 2020, 68, (4), pp 3423–3433. doi: 10.1109/TIE.2020.2982114.DOI
  28. [28]Surve, P., Maity, A. and Kumar, S.R. Polynomial shaping based three-dimensional impact angle and field-of-view constrained guidance, Aerospace Science and Technology, 2024, 147, p 109018. doi: 10.1016/j.ast.2024.109018.DOI
  29. [29]Ghosh, S., Yakimenko, O.A., Davis, D.T. and Chung, T.H. Unmanned aerial vehicle guidance for an all-aspect approach to a stationary point, Journal of Guidance, Control, and Dynamics, 2017, 40, (11), pp 2871–2888. doi: 10.2514/1.G002614.DOI
  30. [30]Kee, P., Dong, L. and Siong, C. Near optimal midcourse guidance law for flight vehicle, 36th AIAA Aerospace Sciences Meeting and Exhibit, 1998, p 583. doi: 10.2514/6.1998-583.DOI

Paper details

vlnc:W_ba2hrscy
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Engineering

Not recorded for this paper

View statement
Looking for related papers…
Contents